The race to capture carbon dioxide (CO2) is on, and researchers at DTU are leading the charge with an innovative approach. Professor Philip Fosbøl is spearheading an EUDP-funded project that leverages artificial intelligence (AI) to revolutionize CO2 capture technology. The goal? To find more efficient and cost-effective solvents for CO2 absorption, a process crucial for mitigating climate change.
The current method involves absorption, where CO2 is absorbed into a solvent, typically made from amines. However, this process is energy-intensive, and the solvents used are not ideal. Here's where AI steps in as a game-changer. By training an AI on vast amounts of data, including published information and computer simulations, it can identify promising solvent candidates from a billion possible combinations.
The AI's approach is akin to a Lego master builder, constructing molecules with nitrogen at their core in various ways. It calculates melting points, heat of vaporization, and intermolecular interactions to suggest the most viable candidates. This AI-driven method is a significant departure from traditional lab-based experimentation, which is time-consuming and resource-intensive.
DTU researchers have already narrowed down the possibilities to 100,000 candidates, carefully selecting two for laboratory testing. This phase is crucial, as it determines the composition and concentrations of the solvent. Randi Neerup, a researcher overseeing the project, explains that the AI doesn't provide a recipe but guides the selection process.
The testing phase, expected to last a year, will involve a total of ten candidates. During this time, the AI will be continuously optimized, incorporating laboratory findings to generate new suggestions. This self-reinforcing process ensures that the technology evolves and adapts based on real-world data.
The implications of this project extend beyond CO2 capture. By developing a versatile AI capable of molecular design for various applications, Hafnium Labs, a project partner, aims to revolutionize industries like cleaning products and food production. This breakthrough could pave the way for more sustainable and efficient processes across multiple sectors.
In conclusion, this DTU project exemplifies how AI can accelerate scientific discovery and innovation. By harnessing the power of AI, researchers are not just capturing CO2 but also capturing the imagination of a more sustainable future. As the project progresses, the world awaits the next breakthrough in CO2 capture technology, one that promises to make a significant impact on our environment.
